Split: Re-scaling experiment

[Zireael]

ShipEntity method - (void) rescaleBy:(GLfloat)factor

That sounds interesting, could it be applied to asteroids and boulders too so that they come in a variety of different sizes?

*jumps up and down* Please say yes!

[Paradox]

Of course humans appear in the game! Many planets claim to be inhabited by them, and, [whisper] don't tell anyone, but I am one too! [/whisper] As a matter of fact, I bet most of us that play Oolite are! (except maybe Smivs...)

Don't worry, at 2m tall this Cucurbitan is very much 'human' size.
And as I explained in a previous post, so are the ships as far as I am concerned. A NASA space shuttle (orbiter) is about 37m long - a Cobra III is a bit bigger at around 65m. All things considered that sounds about right in my book.

Better not tell that to the guys that designed the original Cobra MK3! };]

I do understand your point, Paradox, and even have some sympathy with it, but the fact is, unlike the RL shuttle, many SciFi ships are actually too small to be realistic in the real World (and Oolite) so will never work unless they are 'adapted'.

Ummm Smivs... That's why it's called science fiction... };] And that makes Tie fighters just as legitimately sized as Cobras.

And if the only thing that is needed to get these ships into Oolite is to multiply the measurements by three (which is what many others have done many times before) I really think this is a lot of fuss about nothing.

Let us simply agree to disagree on that point my pumpkinesque friend! };]

[Paradox]

Before changing all ships sizes by hand, maybe it would be worth to investigate the ShipEntity method - (void) rescaleBy:(GLfloat)factor. Right now it is used in the code only for rescaling wreckage meshes, but it might be worth experimenting with it also during new ship spawns, just before adding them to the universe. No guarantees, but it would be easier than re-dimensioning all existing ships. Note that the method takes care of subentities rescaling too.

Interesting idea another_commander. However, that still leaves the ship oxp-ers (as McLane seems to consider this a lesser class of oxp makers...) with the problem of designing a ship, then scaling it up to ridiculous size, just so the game can then scale it back down again... It's just another patch... Granted, after scaling 3 ships now, I am almost tempted, but...

[Zireael]

I say use the rescaling method found by a_c. Because rescaling 99% of the Oolite ships must be a PITA.

There's 17 core ships worth rescaling (I'm not counting non-player ships like Worm and shuttle and Thargoids ) and 242 OXP ships worth rescaling (not counting Killer Wolf's ships and some joke ships).

So while it might be worthwhile to rescale the core 17, all the other ships will get out of whack. And I don't think you want to spend next several years on rescaling.

[Paradox]

I say use the rescaling method found by a_c. Because rescaling 99% of the Oolite ships must be a PITA.

There's 17 core ships worth rescaling (I'm not counting non-player ships like Worm and shuttle and Thargoids ) and 242 OXP ships worth rescaling (not counting Killer Wolf's ships and some joke ships).

So while it might be worthwhile to rescale the core 17, all the other ships will get out of whack. And I don't think you want to spend next several years on rescaling.

You're right... How about this:
A new ship entry, something like "scale_ship = yes", would tell the game engine to scale that particular ship. Adding one line to you're favorite ship's shipdata.plist would not be hard at all. Then "scale_ship = no" means that the ship is already scaled and for the game engine to leave it alone...

This of course assumes that viewpoints and positions etc can be scaled as well...

[Redspear]

However, in the game, what ultimately counts is the player's perspective, under which all other perspectives ultimately have to submit. The (ship) modeller's perspective can therefore not be the defining measure for the game.

Ultimately submit? Must cooperation be to one's detriment?

The different scales in Oolite are not an accident. They are designed to give the player the best possible game experience.

Can we be absoulety certain that they achieve that? Can it be said categorically that the game scales that were built maybe 10 years ago cannot, in any way shape or form, possibly be improved upon with regards to game experience?

This seems to me central to the whole issue.

If you can truly be sure of that and can demonstrate or illustrate it clearly in a way that I can understand then I'd really like to hear it as you'd be saving me a lot of time.

• Changing the scales to realistic ones would likely be to the game's detriment but, as Paradox points out, that's not on the agenda here.
  
  Changing the scales without changing anything else would likely be to the games detriment but that is not being suggested either.
  
  Changing other elements without concern for the current game would likely lead to a very different experience yet I don't think that's been our approach.

So the example you give of launching from the station is a nice one but you're talking about a factor of 100 whilst we're suggesting a factor of 3 (or 3.28 or whatever it is for metres/feet conversion if we really wanted to go there). It's also only one (the largest one at present) of a number of suggestions.

The example you give of the ship at 10,000m is more relevant because now we're talking about the same order of change but wouldn't it depend on what you'd expect to be doing with a ship at that range? Explicitly: weapon ranges, ship speeds and the like would determine this experience which is why they are being discussed. Do you see the ship or do you see the distance? I think the likes of laser and scanner ranges can change your perception in this example and therefore your experience of it.

Yes, there are complications but there are options too. For example: ship speed now too slow? Increase the torus multiplier. Speed can then operate in similar fashion (to the current game) for both combat and for intersystem travel.
Whilst I appreciate that it's not quite that simple (as cim has been very helpful in explaining), I don't think we can say that we have truly encountered any dead ends yet.

On scale in Oolite, from it's creator...

The scale of Oolite was determined by the stations, which are quoted as being 1km in diameter. Then, if the ships stayed at the scales in the manual none of them would have to match the station rotation to fit the slot - they could fly straight in. So I took artistic licence in interpreting the 'ft' of the manual as being equivalent to our metres, the ships then were at the right scale for the docking slit.

From elsewhere on these boards...

All the ships in Oolite are ridiculously large, even when you don’t take the tiny planets into account. Weren’t they at some stage converted from N ft to N m instead of 0.3048N m? If so, the sizes on feet were merely surprising, not silly. :-p

Agreeing here!

The adjustment from feet to meters was necessitated by the increase in scale of the space station to it's full 1km diameter glory. If the ships had stayed scaled by feet then they'd almost all fit in the docking slit sideways - and docking is such an important part of the experience it was necessary to scale ships to match.

The Cobra 3 is particularly huge!

Which was followed by...

Now personally, I’d just have scaled down the docking slit. :-p (Some decor around it, like big arrows pointing inwards and stuff, would stop it from disappearing.)

Somehow, it just didn't look right to me.

On realistic scale...

I know, I'm curious too. The original reason I didn't go full-scale was because the depth-buffer didn't have the granularity for it to work properly, but I worked around that it the end (by employing the old-standby painters' algorithm).

I agree it would be worth the experiment.

and as a caution...

For the moment I'm going to have to say I have too many other things to work on to mess with the scale.

The scale works for current gameplay. A change of scale would change the gameplay and that would lead to a heck of a lot of work. So for the moment, I'll admit it's cartoony, but (as far as my focus goes) it's staying that way for the foreseeable while.

The source is there for anyone who wants to try rescaling. I'd be very interested to see the effects.

That's talking about a full-scale version which is way, way beyond the scope of the experiments proposed on this thread.
The change of gameplay is why speeds, ranges etc. are being discussed in order to at least keep gameplay recognisable.

I aim to test the theory that scale cannot be improved upon without being to the detriment of gameplay.

I think I might just be able to rescale whilst addressing the concerns that you have mentioned and yet I may well fail to do so. One of the big problems with this issue is that there seems to be very little record of anyone trying it and providing details. If nothing else, this thread will become a warning to others of just how problematic such an approach is. That too, it could be argued, would be worthwhile.

The known problems appear navigable at present but, as cim points out, it is the unknown ones that may be my undoing...
If anyone really feels that I'm missing some non-subjective bugbear of an issue (and can explain it) then please point it out.

Thanks.

This thread's getting rather busy. I'll address some of the other stuff later...

Oh, and Paradox... 'Fluffy'? ... Steady on

[Paradox]

However, in the game, what ultimately counts is the player's perspective, under which all other perspectives ultimately have to submit. The (ship) modeller's perspective can therefore not be the defining measure for the game.

Ultimately submit? Must cooperation be to one's detriment?

The different scales in Oolite are not an accident. They are designed to give the player the best possible game experience.

Can we be absoulety certain that they achieve that? Can it be said categorically that the game scales that were built maybe 10 years ago cannot, in any way shape or form, possibly be improved upon with regards to game experience?

This seems to me central to the whole issue.

If you can truly be sure of that and can demonstrate or illustrate it clearly in a way that I can understand then I'd really like to hear it as you'd be saving me a lot of time.

• Changing the scales to realistic ones would likely be to the game's detriment but, as Paradox points out, that's not on the agenda here.
  
  Changing the scales without changing anything else would likely be to the games detriment but that is not being suggested either.
  
  Changing other elements without concern for the current game would likely lead to a very different experience yet I don't think that's been our approach.

So the example you give of launching from the station is a nice one but you're talking about a factor of 100 whilst we're suggesting a factor of 3 (or 3.28 or whatever it is for metres/feet conversion if we really wanted to go there). It's also only one (the largest one at present) of a number of suggestions.

The example you give of the ship at 10,000m is more relevant because now we're talking about the same order of change but wouldn't it depend on what you'd expect to be doing with a ship at that range? Explicitly: weapon ranges, ship speeds and the like would determine this experience which is why they are being discussed. Do you see the ship or do you see the distance? I think the likes of laser and scanner ranges can change your perception in this example and therefore your experience of it.

Yes, there are complications but there are options too. For example: ship speed now too slow? Increase the torus multiplier. Speed can then operate in similar fashion (to the current game) for both combat and for intersystem travel.
Whilst I appreciate that it's not quite that simple (as cim has been very helpful in explaining), I don't think we can say that we have truly encountered any dead ends yet.

On scale in Oolite, from it's creator...

The scale of Oolite was determined by the stations, which are quoted as being 1km in diameter. Then, if the ships stayed at the scales in the manual none of them would have to match the station rotation to fit the slot - they could fly straight in. So I took artistic licence in interpreting the 'ft' of the manual as being equivalent to our metres, the ships then were at the right scale for the docking slit.

From elsewhere on these boards...

All the ships in Oolite are ridiculously large, even when you don’t take the tiny planets into account. Weren’t they at some stage converted from N ft to N m instead of 0.3048N m? If so, the sizes on feet were merely surprising, not silly. :-p

Agreeing here!

The adjustment from feet to meters was necessitated by the increase in scale of the space station to it's full 1km diameter glory. If the ships had stayed scaled by feet then they'd almost all fit in the docking slit sideways - and docking is such an important part of the experience it was necessary to scale ships to match.

The Cobra 3 is particularly huge!

Which was followed by...

Now personally, I’d just have scaled down the docking slit. :-p (Some decor around it, like big arrows pointing inwards and stuff, would stop it from disappearing.)

Somehow, it just didn't look right to me.

On realistic scale...

I know, I'm curious too. The original reason I didn't go full-scale was because the depth-buffer didn't have the granularity for it to work properly, but I worked around that it the end (by employing the old-standby painters' algorithm).

I agree it would be worth the experiment.

and as a caution...

For the moment I'm going to have to say I have too many other things to work on to mess with the scale.

The scale works for current gameplay. A change of scale would change the gameplay and that would lead to a heck of a lot of work. So for the moment, I'll admit it's cartoony, but (as far as my focus goes) it's staying that way for the foreseeable while.

The source is there for anyone who wants to try rescaling. I'd be very interested to see the effects.

That's talking about a full-scale version which is way, way beyond the scope of the experiments proposed on this thread.
The change of gameplay is why speeds, ranges etc. are being discussed in order to at least keep gameplay recognisable.

I aim to test the theory that scale cannot be improved upon without being to the detriment of gameplay.

I think I might just be able to rescale whilst addressing the concerns that you have mentioned and yet I may well fail to do so. One of the big problems with this issue is that there seems to be very little record of anyone trying it and providing details. If nothing else, this thread will become a warning to others of just how problematic such an approach is. That too, it could be argued, would be worthwhile.

The known problems appear navigable at present but, as cim points out, it is the unknown ones that may be my undoing...
If anyone really feels that I'm missing some non-subjective bugbear of an issue (and can explain it) then please point it out.

Thanks.

This thread's getting rather busy. I'll address some of the other stuff later...

Oh, and Paradox... 'Fluffy'? ... Steady on

(I don't know where that came from... it just seemed to fit the moment!)

(Madly dancing around his laptop right now! )

Well, well, well, what an interesting conversation you found there Redspear! The only reason the ships are the size they are, was because of the size of the docking bay! It had nothing to do with seeing 10,000 meters or any of this other... issues. Hey, I bet Griff could make us some killer stations with smaller docking bays if he were so inclined.. };] Not now of course, but eventually maybe...? I can probably re-scale one (just the bay, not the station) but I know there is a bunch of special considerations with bays and stuff that I am not familiar with... Maybe when I get tired of Torchlight...}:] OMG and he called them cartoony as well!

(I gotta get up and dance some more!)

[Diziet Sma]

Smaller docking slots would be cool.. you'd finally get a sense of just how damn big those stations are.. it would be like flying at a wall with a little doorway in it! Freaky!

[cim]

The only reason the ships are the size they are, was because of the size of the docking bay! It had nothing to do with seeing 10,000 meters or any of this other... issues.

Or, as I said a few days ago:

Incidentally, the Elite ships weren't to a consistent scale either: the sizes of the ships are given in feet in the manual, while the station is stated to be 1km across. The relative size of the ships and station in Elite is the same as in Oolite, though. So either the manual is incorrect about station sizes or incorrect about ship sizes. (Or both, of course)

It's hardly the first or most obvious time the Elite manual has stated facts outright contradictory to those shown in-game, of course. And therefore...

1. Were Elite ships originally designed to be a specific size, using the scale of 1 "foot"?

...the answer to question 1 depends on which two of Section 7 (size of Coriolis Station to be measured in Metric units), Section 47 (size of ships to be measured in Imperial units), and in-game experience (relative size of ships and stations to be 6 Cobra-widths to the Coriolis) you hold to be canonical.

It sounds from what you've said that you hold the manual to be correct in both places, and the original Elite game, in providing a visible ratio not in accordance with this, to be heretical. This, arguably, is the stance taken by David Braben too, in that future games in the Elite series have more closely resembled the Elite manual than the Elite game, where the two conflict, so you're in good company there.

In that case I would suggest the following source changes as the simplest approach to correct the ratios:
- adjust the STE range indicator to divide all "metre" ranges by 3.3ish (the base modelling scale is now 1 unit = 1 foot), or alternatively change its unit.
- adjust the initialisation function for StationEntity to rescale the mesh by x3.3 as a default (this will give oversized docking bays but there's nothing short of a complete remodel which will fix that) so that Stations and only Stations are scaled to 1 unit = 1 metre
- adjust the planet scale factor to be x3.3 bigger, and then adjust the planet masslock radius to be x3.3 smaller. Otherwise the stations will be ridiculously large next to the planets.
- play around with the torus drive modifier (x3.3 as well might work, if you sharpened the deceleration)

That means you only need to remodel the stations, rather than all the ships, which is a much shorter job.

OMG and he called them cartoony as well!

I doubt anyone would dispute that the scale - especially the inter-system and planetary scales which are the ones that quote refers to - are cartoony.

If you can truly be sure of that and can demonstrate or illustrate it clearly in a way that I can understand then I'd really like to hear it as you'd be saving me a lot of time.

I would say that the relationship between scale levels that currently exists is at least well-tested, and lots of different components of that scale are integrated into the gameplay. I don't think you can make significant changes to the relationship between "ship scale" and "planetary scale" without also making a significant modification to the way in which the player's ship (and possibly also NPC ships) travel large in-system distances: the key figures if you keep the torus drive being the torus speed multiplier, the masslock radii (scanner range and planetary lock range), and of course the in-system distances themselves. If you switch to a disjoint drive like some of the 8-bit Elites have, or a super-TAF drive like FE2 [1], then that gives more scope to adjust the ratios, but that's a much more major adjustment than a refinement.

[1] Or an alternate-space-scale drive like Elite Dangerous is planned to have, for that matter. That particular solution was necessary to get multiplayer to work but if you're aiming for mostly symmetric NPCs it's also a workable approach for a single-player game.

That sounds interesting, could it be applied to asteroids and boulders too so that they come in a variety of different sizes?

Yes - that's an excellent idea! Relatively straightforward too, I should think, since they don't tend to have subentities, weapons, and all the other things that I don't think that method touches. I'll try it soon.

[cim]

That sounds interesting, could it be applied to asteroids and boulders too so that they come in a variety of different sizes?

Yes - that's an excellent idea! Relatively straightforward too, I should think, since they don't tend to have subentities, weapons, and all the other things that I don't think that method touches. I'll try it soon.

[/quote]The catch, of course, is that mesh rescaling is horribly inefficient: even for a relatively simple asteroid mesh it takes 5-10 times longer to rescale the mesh than the entire rest of the ship addition.

If we want asteroids of different sizes the solution is to provide asteroid models - even if they are just rescales of existing ones - in a variety of different sizes.

[Norby]

My motivation is to address the inter-scale relationships within the game

This is an experiment so here is my ideas about a try to fix more scaling issues in one step.
Ship rescales already need many adjustments in the core so I think similar changes allow a try to improve the planet and maybe the sun scales also.

Ships: I think the scale_ship = yes can be a metadata in ship oxps to flag it is designed to human scale and the core can rescale with the rescaleBy method. For backward compatibility the missing flag mean the oolite ship scale. In the main branch the core can scale up human sized ships, an other (scaled?) branch can scale down all other ships so both can use all ships after some possible bugfixes.

Planets: if we enlarge planets 100 times to get natural sizes then we need much longer lanes so we need much higher torus multipliers and scanner ranges also to avoid missing all other ships without adding 100 times more ships. Interesting challenge but not impossible imho.

Sun and Gas Giants: less problematic if we assume that nobody wants to go around these. In this case the needed 25x further increase over planets maybe thinkable with very extended scooping ranges to made usable these with the previous changes scaled to the planets. Still not end up with the sizes of a full solar system but some really big playable space is imaginable. These huge sizes shouts after Newtonian movements what I am trying to avoid.

make the journey to sunskimming 150.000 times as long as it is now.

It is ok, but 100x is possible within acceptable time if we develop new ideas around long accelerating torus which start braking much sooner scaled with the gained speeds. I plan to make an OXP to confirm this idea.

Inter-system: I think we can keep it as is and accept that these 8 galaxies are some of the denser ones in billions of galaxies.


There is an other problem: the mass of ships. Current vaules much more fit with the smaller ships (in kg) so maybe can stay as is in the human sized branch. This mean mass calculation is need to be changed to get the current values from the smaller volumes.

So I think a private fork named to "real sizes" with smaller ships and 100x sized planets, distances and torus speed is possible after enough works of fans.


On the other hand, I can imagine a new reason in the future to keep the current large ships: if somebody start to build ship interiors which is a question of time imho, for example for detailed damage models or writers use the large inner spaces for human scaled adventures.

The current large Cobra Mk III with its 10 floor height and footbal stadium footprint is contain much larger inner space to make these than the original 3 floor and 4 tennis court scaled one - try to guess how many times (solution is in the next paragraph).

Thanks to Paradox's Scaled CobraMK3 I can tell the difference using the game built-in volume meter named to mass: large Cobra is 185580, but scaled is 5433 only so the large contain 34 times more space for human-sized rooms.


However my current goal is to get 3.3 times more space in the game which is reachable without drastical changes. After enough tests at OXP level what I started to plan, maybe it has a few chance to arrive into the nightly builds in some form imho.

This can improve the gameplay by giving the length of space lanes in Elite and offer more than 10 seconds travels between masslocks.

[Diziet Sma]

The catch, of course, is that mesh rescaling is horribly inefficient: even for a relatively simple asteroid mesh it takes 5-10 times longer to rescale the mesh than the entire rest of the ship addition.

Which would tend to suggest this is not really a practical method for rescaling ships, either..

[another_commander]

The catch, of course, is that mesh rescaling is horribly inefficient: even for a relatively simple asteroid mesh it takes 5-10 times longer to rescale the mesh than the entire rest of the ship addition.

As far as I can recall since the last time I looked at that code (what was it, 1.74? 1.75?) there was another side effect, which is probably still around unless something changed in the meantime: The rescaled mesh was getting written in Oolite's cache, which resulted in the last rescaled models appearing rescaled even when called with the standard spawn functions in subsequent game launches. I believe this is why the wreckage code orders a rescale back to a factor of 1 immediately after the wreckage has been added to the universe.

[cim]

As far as I can recall since the last time I looked at that code (what was it, 1.74? 1.75?) there was another side effect

I fixed that one in 1.77 because calling this sort of rescale on a visual effect was really useful - essential to get sunspots and rings to match their parent object, for instance.

Which would tend to suggest this is not really a practical method for rescaling ships, either.

It wouldn't be so bad there - you'd just rescale the mesh once on first load and stick it into the cache. Not entirely straightforward to code, but it could be efficient.

[Redspear]

That being said... I see them all as numbers, vectors and quaternions. Re-scaling and re-positioning should be just maths.

That sounds like you've got a point there. Thanks

Before changing all ships sizes by hand, maybe it would be worth to investigate the ShipEntity method - (void) rescaleBy:(GLfloat)factor. Right now it is used in the code only for rescaling wreckage meshes, but it might be worth experimenting with it also during new ship spawns, just before adding them to the universe. No guarantees, but it would be easier than re-dimensioning all existing ships. Note that the method takes care of subentities rescaling too.

Ship & subentity models but I'm guessing that it doesn't do player views and scoop/missile/subentity positions???

Hey, if it does anything then it's helpful. Thanks

In that case I would suggest the following source changes as the simplest approach to correct the ratios:
- adjust the STE range indicator to divide all "metre" ranges by 3.3ish (the base modelling scale is now 1 unit = 1 foot), or alternatively change its unit.
- adjust the initialisation function for StationEntity to rescale the mesh by x3.3 as a default (this will give oversized docking bays but there's nothing short of a complete remodel which will fix that) so that Stations and only Stations are scaled to 1 unit = 1 metre
- adjust the planet scale factor to be x3.3 bigger, and then adjust the planet masslock radius to be x3.3 smaller. Otherwise the stations will be ridiculously large next to the planets.
- play around with the torus drive modifier (x3.3 as well might work, if you sharpened the deceleration)

That means you only need to remodel the stations, rather than all the ships, which is a much shorter job.

Thanks cim, that makes sense (but would still leave Paradox with his modelling issue I suspect...)

Is mass-lock from ships purely triggered by their appearance on your scanner?
If so, it might be beneficial for me to increase that a bit (not necessarily by as much as 3.3) in order to approach current encounter rates.

In any case that sounds like a worthy test environment.

I would say that the relationship between scale levels that currently exists is at least well-tested, and lots of different components of that scale are integrated into the gameplay. I don't think you can make significant changes to the relationship between "ship scale" and "planetary scale" without also making a significant modification to the way in which the player's ship (and possibly also NPC ships) travel large in-system distances: the key figures if you keep the torus drive being the torus speed multiplier, the masslock radii (scanner range and planetary lock range), and of course the in-system distances themselves. If you switch to a disjoint drive like some of the 8-bit Elites have, or a super-TAF drive like FE2 [1], then that gives more scope to adjust the ratios, but that's a much more major adjustment than a refinement.

[1] Or an alternate-space-scale drive like Elite Dangerous is planned to have, for that matter. That particular solution was necessary to get multiplayer to work but if you're aiming for mostly symmetric NPCs it's also a workable approach for a single-player game.

Yes, I agree, the current set-up works and many things are adapted to fit it (as you'll know much better than me). This re-scale was always a bit of a long-shot but...

It's nice to hear that the 'key figures' match ones already considered on some level. If adjustments of those can be made in such a way that little else is required then we should be in business.
Replacing the torus-drive with an alternative method is probably beyond my skill at present, so I'll stick to tweaking it for now. Thanks for the ideas though.

My motivation is to address the inter-scale relationships within the game

This is an experiment so here is my ideas about a try to fix more scaling issues in one step.

Hi Norby, glad to have your input

Ships: I think the scale_ship = yes can be a metadata in ship oxps to flag it is designed to human scale and the core can rescale with the rescaleBy method. For backward compatibility the missing flag mean the oolite ship scale. In the main branch the core can scale up human sized ships, an other (scaled?) branch can scale down all other ships so both can use all ships after some possible bugfixes.

Like it, that should make Paradox happy

Planets: if we enlarge planets 100 times to get natural sizes then we need much longer lanes so we need much higher torus multipliers and scanner ranges also to avoid missing all other ships without adding 100 times more ships. Interesting challenge but not impossible imho.

I'd be very interested in anything you achieve in this area but 100 times is so much more ambitious than my current plans. I also wonder if once we get past, say 10 times bigger, will the benefit be significant?

As I said, very interested to hear any ideas on this though...

Sun and Gas Giants: less problematic if we assume that nobody wants to go around these. In this case the needed 25x further increase over planets maybe thinkable with very extended scooping ranges to made usable these with the previous changes scaled to the planets. Still not end up with the sizes of a full solar system but some really big playable space is imaginable. These huge sizes shouts after Newtonian movements what I am trying to avoid.

make the journey to sunskimming 150.000 times as long as it is now.

It is ok, but 100x is possible within acceptable time if we develop new ideas around long accelerating torus which start braking much sooner scaled with the gained speeds. I plan to make an OXP to confirm this idea.

Again, beyond my current ambition but I don't doubt that you're a much more experienced coder than me so please let me know your findings

I used to wonder if the torus speed could change depending on say your distance from the witch-point?
If the further you got from the witch-point the faster it went then longer distances might be navigable.
If the gradient of change was a gradual one then it might appear subtle in game.
There might also be a more realistic effect of planets appearing small for a long time and then more abrubtly starting to get bigger as you begin to get close to them (I explained that quite badly but I hope it makes sense).
The decelleration on the torus would have to be very strong in that situation I suspect.
Anyway, that's for another time perhaps...

So I think a private fork named to "real sizes" with smaller ships and 100x sized planets, distances and torus speed is possible after enough works of fans.

Maybe one day but let's get 3.3 working first, if we can...

However my current goal is to get 3.3 times more space in the game which is reachable without drastical changes. After enough tests at OXP level what I started to plan, maybe it has a few chance to arrive into the nightly builds in some form imho.

If everyone loves it then sure, why not? but I don't think that's likely to happen any time soon.
I think we need to remember that any experiment is just that. If we can make it truly great then like you say, maybe it has a chance.
But first, work to do, right?

As you say, see if we can get 3.3 to work and then if we can do that there will probably be some more adjusting to do to get it to work better, and perhaps closer (gameplay wise) to the current Oolite.
e.g. If we have to increase scanner range even a little then that could change the start of every encounter with other ships. Something to ponder...

Thanks again for your interest and support